| Title: | StreamScan: Fast scan algorithms for GPUs without global barrier synchronization |
| Author: | Yan, Shengen (1)
; Long, Guoping (1)
; Zhang, Yunquan (1)
|
| Conference Name: | 18th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2013
|
| Conference Date: | February 23, 2013 - February 27, 2013
|
| Issued Date: | 2013
|
| Conference Place: | Shenzhen, China
|
| Keyword: | Scan
; prefix-sum
; OpenCL
; CUDA
; GPU
; Parallel algorithms
|
| Corresponding Author: | Yan, S.(yanshengen@gmail.com)
|
| Publish Place: | Association for Computing Machinery, General Post Office, P.O. Box 30777, NY 10087-0777, United States
|
| Indexed Type: | SCI
; EI
|
| ISSN: | 0362-1340
|
| ISBN: | 9781450319225
|
| Department: | (1) Lab. of Parallel Software and Computational Science, Institute of Software, Chinese Academy of Sciences, Beijing, China; (2) State Key Laboratory of Computing Science, Chinese Academy of Sciences, Beijing, China; (3) Graduate University, Chinese Academy of Sciences, Beijing, China
|
| Abstract: | Scan (also known as prefix sum) is a very useful primitive for various important parallel algorithms, such as sort, BFS, SpMV, compaction and so on. Current state of the art of GPU based scan implementation consists of three consecutive Reduce-Scan-Scan phases. This approach requires at least two global barriers and 3N (N is the problem size) global memory accesses. In this paper we propose StreamScan, a novel approach to implement scan on GPUs with only one computation phase. The main idea is to restrict synchronization to only adjacent workgroups, and thereby eliminating global barrier synchronization completely. The new approach requires only 2N global memory accesses and just one kernel invocation. On top of this we propose two important op-timizations to further boost performance speedups, namely thread grouping to eliminate unnecessary local barriers, and register optimization to expand the on chip problem size. We designed an auto-tuning framework to search the parameter space automatically to generate highly optimized codes for both AMD and Nvidia GPUs. We implemented our technique with OpenCL. Compared with previous fast scan implementations, experimental results not only show promising performance speedups, but also reveal dramatic different optimization tradeoffs between Nvidia and AMD GPU platforms. © 2013 ACM. |
| English Abstract: | Scan (also known as prefix sum) is a very useful primitive for various important parallel algorithms, such as sort, BFS, SpMV, compaction and so on. Current state of the art of GPU based scan implementation consists of three consecutive Reduce-Scan-Scan phases. This approach requires at least two global barriers and 3N (N is the problem size) global memory accesses. In this paper we propose StreamScan, a novel approach to implement scan on GPUs with only one computation phase. The main idea is to restrict synchronization to only adjacent workgroups, and thereby eliminating global barrier synchronization completely. The new approach requires only 2N global memory accesses and just one kernel invocation. On top of this we propose two important op-timizations to further boost performance speedups, namely thread grouping to eliminate unnecessary local barriers, and register optimization to expand the on chip problem size. We designed an auto-tuning framework to search the parameter space automatically to generate highly optimized codes for both AMD and Nvidia GPUs. We implemented our technique with OpenCL. Compared with previous fast scan implementations, experimental results not only show promising performance speedups, but also reveal dramatic different optimization tradeoffs between Nvidia and AMD GPU platforms. © 2013 ACM. |
| Language: | 英语
|
| Citation statistics: |
|
| Content Type: | 会议论文
|
| URI: | http://ir.iscas.ac.cn/handle/311060/16554
|
| Appears in Collections: | 软件所图书馆_会议论文
|
|
There are no files associated with this item.
|
| Recommended Citation: | |
Yan, Shengen ,Long, Guoping ,Zhang, Yunquan . StreamScan: Fast scan algorithms for GPUs without global barrier synchronization[C]. 见:18th ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming, PPoPP 2013. Shenzhen, China. February 23, 2013 - February 27, 2013.
|
|
|
